THE ROLE OF EXPECTATIONS ON WATERFOWL HUNTER SATISFACTION. Kevin D. Brunke

Transcription

1 THE ROLE OF EXPECTATIONS ON WATERFOWL HUNTER SATISFACTION By Kevin D. Brunke A Thesis Submitted to the Faculty of Mississippi State University in Partial Fulfillment of the Requirements for the Degree of Master of Science in Wildlife and Fisheries Science in the Department of Wildlife and Fisheries Mississippi State, Mississippi May 2007

2 THE ROLE OF EXPECTATIONS ON WATERFOWL HUNTER SATISFACTION BY Kevin D. Brunke Approved: LA.& Kevin M. Hunt, Ph.D. Assistant Professor of Wildlife and Fisheries (Director of Thesis), '~rofessor of ~ o h t r ~ (Committee Member) Professor of Wildlife and Fisheries (Committee Member) Bruce D. Leouold, ~kr'fl. Professor of wildlife and Fisheries (Department HeadJGraduate Coordinator)

3 Name: Kevin D. Brunke Date of Degree: May 4, 2007 Institution: Mississippi State University Major Field: Wildlife and Fisheries Science Major Professor: Dr. Kevin M. Hunt Title of Study: THE ROLE OF EXPECTATIONS ON WATERFOWL HUNTER SATISFACTION Pages in Study: 59 Candidate for Degree of Master of Science Hunter satisfaction has received extensive attention in the literature, but the role of expectations on satisfaction has been neglected. Consumer satisfaction researchers often use the expectancy disconfirmation paradigm (i.e., differences between expectations and reality) to address relationships between expectations and satisfaction. I used this paradigm to examine the relationship between expectations and satisfaction for waterfowl hunters in Arkansas and Mississippi. I found hunter satisfaction was a partial function of fulfilled expectations in both studies. Performance-only measures generally correlated more strongly with overall satisfaction than disconfirmations measured by a difference score. Conversely, disconfirmation of expectations for a season measured on a single item scale, had the greatest relationship with overall satisfaction for a season. Knowledge of congruence between hunter expectations and outcomes offers managers an avenue to effectively focus management efforts to improve satisfaction levels.

4 DEDICATION I dedicate this thesis to Casey, my loving wife. Thank you for giving me love and encouragement while pursuing my dreams. ii

5 ACKNOWLEDGEMENTS I want to express my sincerest appreciation to those who assisted in the completion of this thesis. First of all, I thank Dr. Kevin M. Hunt, committee chairman, for his guidance and assistance throughout my Master of Science program. I also want to express appreciation to my committee members, Dr. Richard M. Kaminski and Dr. Stephen C. Grado, for their input on my thesis and other projects during my tenure at Mississippi State University. I also thank my funding sources: U. S. Department of the Interior - U. S. Fish and Wildlife Service, Natural Resource Enterprises and the Mississippi Department of Wildlife, Fisheries, and Parks (MDWFP Project W-48-54, Job 6). I also would like to thank the staff at the Monsanto Farm and Wildlife Management Center for their help with implementing my study and for distributing questionnaires at the center, especially Mr. Ray Bohanan, Mr. Shane Roethle, and Dr. John Anderson. Lastly, I would like to thank my fellow graduate students and everyone in the Human Dimension and Conservation Law Enforcement Laboratory for their diligent efforts during survey design, survey mailing, data entry, and data analyses. Specifically, I thank Edith Fogarty, Corey Wigginton, Nanda Joginipally, Nathan Gregory, Brittany Harris, Elizabeth Kirk, Morgan Miranda, Katie Nelson, and Mallory Sullivan. iii

6 TABLE OF CONTENTS Page DEDICATION... ACKNOWLEDGEMENTS... LIST OF TABLES... LIST OF FIGURES... ii iii v vii CHAPTER I. INTRODUCTION AND OBJECTIVES... 1 Introduction... 1 Objectives... 6 References... 6 II. III. COMPARISON OF TWO APPROACHES FOR THE MEASUREMENT OF WATERFOWL HUNTER SATISFACTION Introduction Methods Results Discussion and Implications References MISSISSIPPI WATERFOWL HUNTER EXPECTATIONS, SATISFACTION, AND INTENTIONS TO HUNT IN THE FUTURE Introduction Methods Results Discussion and Implications References iv

7 LIST OF TABLES TABLE Page 2.1 Means for item expectations, performance, and difference scores for a two-day waterfowl hunting trip near Stuttgart, Arkansas during the waterfowl hunting season. Items were classified as positive disconfirmation, confirmation, or negative disconfirmation based on the 95% confidence intervals around the mean difference score (n = 143) The results of performance-only items correlated with overall satisfaction for a two-day waterfowl hunting trip near Stuttgart, Arkansas during the waterfowl hunting season; ranked by Spearman s rho (n = 71) The results of difference scores correlated with overall satisfaction for a two-day waterfowl hunting trip near Stuttgart, Arkansas during the waterfowl hunting season; ranked by Spearman s rho (n = 70) Differences between difference scores (n = 70) and performance-only (n = 71) correlations with overall satisfaction for a two-day waterfowl hunting trip near Stuttgart, Arkansas during the waterfowl hunting season. Differences were detected using Fisher s z-test and items are ranked by p-value Frequencies and means for items used in the backward step-wise hierarchical log-linear analysis on Mississippi waterfowl hunters (n = 566) Backward step-wise hierarchical log-linear analysis output for Mississippi waterfowl hunters for the model including a 60-day duck season with a six-duck limit (n = 545) Backward step-wise hierarchical log-linear analysis output for Mississippi waterfowl hunters for the model including a 45-day duck season with a six-duck limit (n = 539) v

8 TABLE Page 3.4 Backward step-wise hierarchical log-linear analysis output for Mississippi waterfowl hunters for the model including a 30-day duck season with a three-duck limit (n = 537) vi

9 LIST OF FIGURES FIGURE Page 3.1 An illustration demonstrating the relationships for the backward step-wise hierarchical log-linear model for Mississippi waterfowl hunters, including a 60-day duck season with a six-duck limit. A line between items indicates a significant relationship between items in the final model (n = 545) An illustration demonstrating the relationships for the backward step-wise hierarchical log-linear model for Mississippi waterfowl hunters, including a 45-day duck season with a six-duck limit. A line between items indicates a significant relationship between items in the final model (n = 539) An illustration demonstrating the relationships for the backward step-wise hierarchical log-linear model for Mississippi waterfowl hunters, including a 30-day duck season with a three-duck limit. A line between items indicates a significant relationship between items in the final model (n = 537) vii

10 CHAPTER I INTRODUCTION AND OBJECTIVES Introduction Three theories have been used to describe hunter satisfaction and benefits. The game-bagged theory was first to emerge, but research revealed that bagging game was not the sole motivation for most hunters (Hendee, 1974). Crissey (1971) suggested a days-afield approach, where number of days spent in the field was considered a suitable measure of hunter benefits. However, this measure left out hunting quality and hunter satisfaction and was short-lived (Hendee, 1974). Hendee (1974) was among the first to suggest a multiple-satisfaction approach to game management, a theory supported widely in the literature (Vaske, Fedler, & Graefe, 1986; Hammitt, McDonald, & Patterson, 1990; Hazel, Langenau, & Levine, 1990; Frey, Conover, Borgo, & Messmer, 2003). This concept asserts there is more to hunter satisfaction than bagging game and days-afield, and recreational resources should be managed to offer people opportunity for a range of experiences (Hendee, 1974). Whereas much research supports the multiple satisfaction approach, the greatest antecedents to hunter satisfaction have varied among studies. Some researchers have found success-related factors (i.e., harvest or getting shots) were strong predictors of satisfaction (Stankey, Lucas, & Ream, 1973; Decker, Brown, & Gutierrez, 1980; 1

11 2 Gigliotti, 2000), while others have reported non-success aspects of the hunt (i.e., getting outdoors or getting close to nature) were strong predictors of satisfaction (Hammitt et al., 1990; Hayslette, Armstrong, & Mirarchi, 2001). Generally, studies have found harvestoriented hunters were less satisfied than hunters motivated by other reasons (Decker et al., 1980; Gigliotti, 2000). Other studies have shown successful hunters reported greater levels of satisfaction than unsuccessful ones (Vaske, Donnelly, Heberlein, & Shelby, 1982; Gigliotti, 2000). Nevertheless, many researchers have indicated the importance of maintaining some probability of harvest success to uphold hunter satisfaction (Stankey et al., 1973; Decker et al., 1980; McCullough & Carmen, 1982). Thus, hunter satisfaction is a complex issue, and no standardized set of measures has been developed to predict it universally. When satisfaction is examined among other disciplines, research suggests the true construct is conceptually rooted in expectancy theory (Manning, 1999). Ultimately, satisfaction is defined as the congruence between expectations and outcomes (Manning, 1999). Although the concept of satisfaction is rooted in expectancy theory and numerous definitions of satisfaction relate to expectations, the role of expectations on satisfaction has been largely neglected in hunter-satisfaction literature. Researchers have compared actual hunting experiences to an ideal hunt in some studies (e.g., Decker et al., 1980), but the probability of a hunter experiencing an ideal hunt with consistency is probably quite low. Other authors have found anglers with realistic expectations for fish size had greater fishing and trip satisfaction than anglers with unrealistic expectations (Spencer & Spangler, 1992). However, the relationship between expectations and satisfaction has

12 been studied extensively by consumer-satisfaction researchers under the expectancy 3 disconfirmation paradigm (Oliver, 1980; Tse & Wilton, 1988; Van Ryzin, 2004). Using this paradigm, individuals compare actual performance to standards of performance they formulated based on their expectations (Niedrich, Kiryanova, & Black, 2005). Accordingly, disconfirmation occurs when a difference exists between one s expectations and the reality of the experience (Oliver & DeSarbo, 1988; Burns, Graefe, & Absher, 2003). Negative disconfirmation occurs when reality is worse than expectations, positive disconfirmation occurs when reality is better than expectations, and confirmation results when reality and expectations are similar (Oliver, 1980). As one progresses from negative to positive disconfirmation, ratings of satisfaction generally increase (Tse & Wilton, 1988; Spreng, MacKenzie, & Olshavsky, 1996). Disconfirmations are typically formed from three expectancies that can be generalized as goals, normative expectations, and predictive expectations (Boulding, Kalra, Staelin, & Zeithaml, 1993; Niedrich et al., 2005). First, goals may be classified as something someone wants or needs and are less susceptible to change by marketing than other types of expectancies (Boulding et al., 1993; Niedrich et al., 2005). Goals are similar to motivations studied by hunter satisfaction researchers in that goals are the starting point of willful control of action to fulfill a want or need (Niedrich et al., 2005). Thus, fulfilling a person s goals may affect their satisfaction in the same way fulfilled motivations have been found to by hunter satisfaction researchers. Second, normative expectations are something someone thinks should happen during their next encounter with an event (Boulding et al., 1993; Niedrich et al., 2005). Finally, predictive

13 expectations are something someone thinks will happen during their next experience 4 with an event and are the standard of comparison typically used when studying consumer satisfaction (Boulding et al., 1993; Niedrich et al., 2005). Furthermore, expectancies are updated with each encounter with an event and change over time based on perceptions of performance, situational influences, or personal characteristics (Boulding et al., 1993; Niedrich et al., 2005). Disconfirmation of expectancies has been measured by a subjective evaluation of how well expectations were met on a post-experience measurement scale (Oliver, 1980), or by a difference score derived by taking the perception of performance minus the expectation of performance (Parasuraman, Zeithaml, & Berry, 1988). However, much debate exists on the use of difference scores versus performance-only measures as predictors of satisfaction and service quality (Cronin & Taylor, 1992; Cronin & Taylor, 1994; Parasuraman, Zeithaml, & Berry, 1994; Burns et al., 2003). Performance-only measures have been found to be better predictors of overall satisfaction ratings and perceived service quality than difference scores in some studies (Cronin & Taylor, 1992; Burns et al., 2003). However, difference scores are useful at identifying deficient areas and tracking expectations over time (Parasuraman et al., 1994; Crompton & Love, 1995). Burns et al. (2003) indicated difference scores may become better predictors of overall satisfaction if measurement of visitor expectations occurred prior to the recreation experience. Others have found different types of comparison standards form separate constructs when measured as an expectation, but when want, need, will, and should comparison standards are measured as disconfirmations on a post-experience

14 5 scale, the standards acted interchangeably (Niedrich et al., 2005). Niedrich et al. (2005) concluded that researchers should decide if the potential loss of reliability in difference scores is worse than the potential loss of construct dimensionality in post-experience measurements of disconfirmation. Performance-only measures are typical of predictor variables used by hunter-satisfaction researchers. Developing a sound understanding of hunter satisfaction is important because there has been a downward trend in hunting participation across much of the United States (Enck, Decker, & Brown, 2000; Li, Zinn, Barro, & Manfredo, 2003; Mehmood, Zhang, & Armstrong, 2003). Many constraints to hunter participation have been found to be strong predictors of satisfaction (Barro & Manfredo, 1996; Miller & Vaske, 2003; Fulton & Hundertmark, 2004; Fulton & Manfredo, 2004). Additionally, satisfaction with an experience has been shown to increase future intentions for participating in an activity (Tian-Cole, Crompton, & Willson, 2001). Moreover, strong predictors of hunter satisfaction, such as harvest and game abundance, have been demonstrated to increase participation among waterfowl hunters (Miller & Hay, 1981; Ringelman, 1997). Thus, expectations, satisfactions, and behavioral intentions are interrelated among hunters, and research examining these relationships is warranted if managers desire the biological, political, and economic support from hunters (Enck et al., 2000; Grado, Kaminski, Munn, & Tullos, 2001). My research was designed to increase understanding of waterfowl hunter expectations, satisfactions, and behavioral intentions.

15 Objectives 6 The primary objectives of my thesis were to: 1) Determine if service quality shortfalls existed on a privately owned waterfowl hunting property. 2) Determine if performance-only scores correlated better with overall satisfaction than difference scores. 3) Determine the greatest antecedents for waterfowl hunter satisfaction. 4) Determine the relationship among expectations, satisfaction, and future behavioral intentions of waterfowl hunters. References Barro, S. C., & Manfedo, M. J. (1996). Constraints, psychological investment, and hunting participation: Development and testing of a model. Human Dimensions of Wildlife, 1, Boulding, W., Kalra, A., Staelin, R., & Zeithaml, V. A. (1993). A dynamic process model of service quality: From expectations to behavioral intentions. Journal of Marketing Research, 30, Burns, R. C., Graefe, A. R., & Absher, J. D. (2003). Alternate measurement approaches to recreational customer satisfaction: Satisfaction-only versus gap scores. Leisure Sciences, 25, Crissey, W. F. (1971). Some thoughts on wildlife research and management objectives. Wildlife Society News, 134, Crompton, J. L., & Love, L. L. (1995). The predictive validity of alternative approaches to evaluating quality of a festival. Journal of Travel Research, 34, Cronin, J. J., Jr., & Taylor, S. A. (1992). Measuring service quality: A reexamination and extension. Journal of Marketing, 56,

16 Cronin, J. J., Jr., & Taylor, S. A. (1994). SERVPERF versus SERVQUAL: Reconciling performance-based and perceptions-minus-expectations measurement of service quality. Journal of Marketing, 58, Decker, D. J., Brown, T. J., & Gutierrez, R. J. (1980). Further insights into the multiplesatisfactions approach for hunter management. Wildlife Society Bulletin, 8, Enck, J. W., Decker, D. J., and Brown, T. L. (2000). Status of hunter recruitment and retention in the United States. Wildlife Society Bulletin, 28, Frey, S. N., Conover, M. R., Borgo, J. S., & Messmer, T. A. (2003). Factors influencing pheasant hunter harvest and satisfaction. Human Dimensions of Wildlife, 8, Fulton, D. C., & Hundertmark, K. (2004). Assessing the effects of a selective harvest system on moose hunters' behaviors, beliefs, and satisfaction. Human Dimensions of Wildlife, 9, Fulton, D. C., & Manfredo, M. J. (2004). A panel design to assess the effects of regulatory induced reductions in opportunity on deer hunters' satisfaction. Human Dimensions of Wildlife, 9, Gigliotti, L. M. (2000). A classification scheme to better understand satisfaction of Black Hills deer hunters: The role of harvest success. Human Dimensions of Wildlife, 5, Grado, S. C., Kaminski, R. M., Munn, I. A., & Tullos, T. A. (2001). Economic impacts of waterfowl hunting on public lands and at private lodges in the Mississippi Delta. Wildlife Society Bulletin, 29, Hammitt, W. E., McDonald, C. D., & Patterson, M. E. (1990). Determinants of multiple satisfaction for deer hunting. Wildlife Society Bulletin, 18, Hayslette, S. E., Armstrong, J. B., & Mirarchi, R. E. (2001). Mourning dove hunting in Alabama: Motivations, satisfactions, and sociocultural influences. Human Dimensions of Wildlife, 6, Hazel, K. L., Langenau, E. E., & Levine, R. L. (1990). Dimensions of hunting satisfaction: Multiple-satisfactions of wild turkey hunting. Leisure Sciences, 12, Hendee, J. C. (1974). A multiple-satisfaction approach to game management. Wildlife Society Bulletin, 2,

17 Li, C.-L., Zinn, H. C., Barro, S. C., & Manfredo, M. J. (2003). A cross-regional comparison of recreation patterns of older hunters. Leisure Sciences, 25, Manning, R. E. (1999). Studies in outdoor recreation: Search and research for satisfaction (2nd ed.). Corvallis, OR: Oregon State University Press. McCullough, D. R., & Carmen, W. J. (1982). Management goals for deer hunter satisfaction. Wildlife Society Bulletin, 10, Mehmood, S., Zhang, D., & Armstrong, J. (2003). Factors associated with declining hunting license sales in Alabama. Human Dimensions of Wildlife, 8, Miller, C. A., & Vaske, J. J. (2003). Individual and situational influences on declining hunter effort in Illinois. Human Dimensions of Wildlife, 8, Miller, J. R., & Hay, M. J. (1981). Determinants of hunter participation: Duck hunting in the Mississippi Flyway. American Agricultural Economics Association, 64, Niedrich, R. W., Kiryanova, E., & Black, W. C. (2005). The dimensional stability of the standards used in the disconfirmation paradigm. Journal of Retailing, 81, Oliver, R. L. (1980). A cognitive model of the antecedents and consequences of satisfaction decisions. Journal of Marketing Research, 17, Oliver, R. L., & DeSarbo, W. S. (1988). Response determinants in satisfaction judgments. Journal of consumer research, 14, Parasuraman, A., Zeithaml, V. A., & Berry, L. L. (1988). SERVQUAL: A multiple-item scale for measuring consumer perceptions of service quality. Journal of Retailing, 64, Parasuraman, A., Zeithaml, V. A., & Berry, L. L. (1994). Reassessment of expectations as a comparison standard in measuring service quality: Implications for further research. Journal of Marketing, 58, Ringelman, J. K. (1997). Effects of regulations and duck abundance on duck hunter participation and satisfaction. Transactions of the North American Wildlife and Natural Resource Conference, 62, Spencer, P. D., & Spangler, G. R. (1992). Effect that providing fishing information has on angler expectations and satisfaction. North American Journal of Fisheries Management, 12,

18 Spreng, R. A., MacKenzie, S. B., & Olshavsky, R. W. (1996). A reexamination of the determinants of consumer satisfaction. Journal of Marketing, 60, Stankey, G. H., Lucas, R. C., & Ream, R. R. (1973). Relationships between hunting success and satisfaction. Transactions of the North American Wildlife Conference, 38, Tian-Cole, S., Crompton, J. L., & Willson, V. L. (2002). An empirical investigation of the relationships between service quality, satisfaction and behavioral intentions among visitors to a wildlife refuge. Journal of Leisure Research, 34, Tse, D. K., & Wilton, P. C. (1988). Models of consumer satisfaction formation: An extension. Journal of Marketing Research, 25, Van Ryzin, G. G. (2004). Expectations, performance, and citizen satisfaction with urban services. Journal of Policy Analysis and Management, 23, Vaske, J. J., Donnelly, M. P., Heberlein, T. A., & Shelby, B. (1982). Differences in reported satisfaction ratings by consumptive and nonconsumptive recreationists. Journal of Leisure Research, 14, Vaske, J. J., Fedler, A. J., & Graefe, A. R. (1986). Multiple determinants of satisfaction from a specific waterfowl hunting trip. Leisure Sciences, 8,

19 CHAPTER II COMPARISON OF TWO APPROACHES FOR THE MEASUREMENT OF WATERFOWL HUNTER SATISFACTION Introduction Hunter satisfaction has been researched extensively, but the strongest antecedents to satisfaction have varied among studies. Some researchers reported success-related factors (i.e., harvest, seeing harvestable wildlife, or getting shots) were strong predictors of satisfaction (Stankey, Lucas, & Ream, 1973; Decker, Brown, & Gutierrez, 1980; Gigliotti, 2000), whereas others reported non-success aspects of the hunt (i.e., getting outdoors or close to nature) were greatly correlated with satisfaction (Hammitt, McDonald, & Patterson, 1990; Hayslette, Armstrong, & Mirarchi, 2001). Others have shown successful hunters reported greater levels of satisfaction than unsuccessful ones (Vaske, Donnelly, Heberlein, & Shelby, 1982; Gigliotti, 2000). Although hunter satisfaction is more than simply harvesting animals (Hendee, 1974), the opportunity to harvest plays an important role in satisfaction judgments (Decker et al., 1980; Gigliotti, 2000; Miller & Graefe, 2001). However, harvest often is a goal hunters have the least control of achieving and may subsequently report lesser levels of satisfaction than participants in other types of recreational activities (Vaske et al., 1982). Thus, hunter satisfaction is viewed as a multifaceted concept depending on the participant and 10

20 11 recreational setting (Hazel, Langenau, & Levine, 1990; Manning, 1999; Frey, Conover, Borgo, & Messmer, 2003). Previous research on recreational satisfaction indicates satisfaction is the difference between desired outcomes and perceived fulfillment of those outcomes (Lawler, 1973; Ditton, Graefe, & Fedler, 1981; Holland & Ditton, 1992). Given this definition, hunters may formulate expectations for the fulfillment of desired outcomes before their hunting experiences. Previously, researchers also have suggested expectations may play a role in hunter satisfaction (Vaske, Fedler, & Graefe, 1986; Hammitt et al., 1990), or there was need to provide hunter education to bring expectations closer to reality (Decker et al., 1980; Gigliotti, 2000). However, research on the relationship between hunter expectations and satisfaction is lacking. The dominant approach for addressing similar relationships elsewhere is in the expectancy disconfirmation paradigm, which has been studied extensively in consumer satisfaction research (Oliver, 1980; Tse & Wilton, 1988; Van Ryzin, 2004). In the expectancy disconfirmation paradigm, disconfirmation occurs when a difference exists between one s expectations and the reality of the experience (Oliver & DeSarbo, 1988; Burns, Graefe, & Absher, 2003). Negative disconfirmation occurs when reality is worse than expectations, positive disconfirmation occurs when reality is better than expectations, and confirmation occurs when reality and expectations are similar (Oliver, 1980). When positive disconfirmation occurs, satisfaction ratings should be greater than when negative disconfirmation occurs. Literature on consumer satisfaction

21 12 has shown the positive relationship between disconfirmation and satisfaction generally has held true (Tse & Wilton, 1988; Spreng, MacKenzie, & Olshavsky, 1996). Disconfirmation has either been measured by a subjective evaluation of the fulfillment of expectations on a post-experience measurement scale (Oliver, 1980) or by a difference score derived by taking the perception of performance minus the expectation of performance (Parasuraman, Zeithaml, & Berry, 1988). Much debate exists on use of difference scores versus performance-only measures as predictors of satisfaction and service quality (Cronin & Taylor, 1992; Cronin & Taylor, 1994; Parasuraman, Zeithaml, & Berry, 1994; Burns et al., 2003). Performance-only measures were better measures of overall satisfaction ratings and perceived service quality than difference scores in some studies (Cronin & Taylor, 1992; Burns et al., 2003). However, Burns et al. (2003) suggested difference scores may have correlated better with overall satisfaction if a different type of expectation was measured or the measurement of expectations occurred before the recreational experience. Further, difference scores may be better at identifying deficient areas and tracking expectations over time than performance-only measures (Parasuraman et al., 1994; Crompton & Love, 1995). Others concluded measuring disconfirmations with a difference score may reduce reliability; however, disconfirmations measured on a post-experience measurement scale may cause loss of construct dimensionality (Niedrich, Kiryanova, & Black, 2005). Furthermore, performance-only measures of satisfaction do not allow for any type of measurement of disconfirmation and have been the typical independent variables for hunter satisfaction researchers.

22 Several types of expectations have been defined and studied in expectancy 13 disconfirmation research, but the standards of predictive and normative expectations emerge as dominant (Boulding, Kalra, Staelin, & Zeithaml, 1993; Niedrich et al., 2005). Predictive expectations (i.e., expectations that people believe will happen) are futureoriented beliefs about uncertain outcomes generated from participants personal experiences, communication with other people, and related beliefs (Boulding et al., 1993; Niedrich et al., 2005). Normative expectations (i.e., expectations that people believe should happen) typically include a larger set of attribute information than predictive expectations and comprise outcomes people believe they deserve (Boulding et al., 1993; Niedrich et al., 2005). Boulding et al. (1993) also indicated predictive and normative expectations may change over time with additional contacts or experiences. Predictive expectations (i.e., will expectations) have typically been used in satisfaction research, whereas normative expectations (i.e., should expectations) have been used in service quality research (Boulding et al., 1993). I used difference scores (i.e., performance expectations) and performance-only measures to examine satisfaction among waterfowl hunters at a private waterfowl hunting area in Arkansas. Assuming harvest is a goal hunters have least control in achieving (Vaske et al., 1982), I hypothesized negative disconfirmation would be greatest for harvest-related variables. Additionally, I hypothesized harvest-related variables would be significant predictors of overall satisfaction. Based on previous research (e.g., Burns et al., 2003), I hypothesized performance-only measures would correlate better with overall

23 satisfaction than difference scores. Finally, I hypothesized satisfaction would be 14 influenced positively by fulfillment of expectations. Methods My study site was a 1,214 hectare farm and hunting lodge owned by Monsanto Company with approximately 384 hectares of artificially floodable forest land. The site is located less than eight kilometers south of Stuttgart, Arkansas. Ducks were primarily hunted in the flooded forest land. Monsanto Company staff invited waterfowl hunters to the property for a two-day hunting trip with lodging, meals, entertainment, and guides provided on-site. Guests were encouraged to bring their own firearms, ammunition, hunting clothes, and personal items. Arkansas hunting licenses and federal waterfowl stamps were available on-site for those who did not already have these. Guests typically arrived the evening before the first day s hunt. After dinner, hunting licenses were processed and a mandatory safety video was shown to guests. Staff distributed a voluntary 4-page pre-hunt questionnaire, with an informational flyer, among guests while their licenses were processed. In the pre-hunt questionnaire, I asked participants to provide their name and address, and using a five-point Likert type scale with the response format of 1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, and 5 = strongly agree, rate how strongly they agreed or disagreed with statements about their expectations. Expectations were related to three aspects of their trip to the property: 1) hunting experience, 2) service from the staff, and 3) impressions of the facilities. Following Niedrich et al. (2005) and Boulding et al. (1993), I operationalized these predictive expectations with the word will for each predictive statement. After

24 participants completed the pre-hunt questionnaire, participants placed finished 15 questionnaires in a locked ballot box until retrieved by researchers. I mailed an eightpage post-hunt questionnaire to participants who completed the pre-hunt questionnaire one week after the duck hunting season to quantify the fulfillment of expectations during their trip. I used techniques modified from Dillman (1978) for conducting mail surveys. I mailed the first post-hunt questionnaires with a cover letter and a postage-paid business reply envelope (hereafter termed complete packet). Then, I directed a second complete packet to non-respondents three weeks after initial contact. After the second contact, I mailed non-respondents a final complete packet four weeks after the previous mailing. I included all questionnaires received within six weeks of the final mailing in data analyses. I pre-tested both survey instruments during the waterfowl hunting season on 94 individuals who hunted the Monsanto property. Based on the pre-test, I reworded some questionnaire items to alleviate ambiguity and reorganized items to lessen item non-response. I reworded the post-hunt questionnaire items from the pre-hunt questionnaire to gauge guest expectation fulfillment via difference scores. I used the same response format for the pre- and post-hunt questionnaires. The post-hunt questionnaire also had questions on harvest, waterfowl hunting participation, satisfaction, trip expenditures, preferences for management of the property, and hunter demographics. I asked hunters to rate their satisfactions with their waterfowl hunting trip using a five-point satisfaction continuum with a response format of 1 = not at all satisfied, 2 = slightly satisfied, 3 =

25 16 moderately satisfied, 4 = very satisfied, and 5 = extremely satisfied. Satisfaction items were related to: 1) their hunting experience, 2) service from the staff, 3) their impressions of the facilities, and 4) the trip overall. After I received questionnaires, I coded, entered, verified, and analyzed the data. I calculated difference scores based on Parasuraman et al. s (1988) work on service quality. I derived difference scores by subtracting participants performance ratings from their expectations ratings for each item related to their trip. Difference scores with negative values indicated participants expectations were not met and negative disconfirmation occurred. Unlike Parasuraman et al. (1988), I used 95% confidence intervals to determine if scores varied significantly from zero and then assigned an item to either positive disconfirmation, confirmation, or negative disconfirmation. I measured the relative importance of each item to overall satisfaction using a one-tailed Spearman s rho, because my hypotheses were directional and the data being analyzed were ordinal (Schlotzhauer & Littell, 1997). To determine if performance-only scores correlated better with overall satisfaction than difference scores, I used methods modified from Burns et al. (2003). I split data from respondents randomly into two halves to meet the assumption of independence for subsequent analysis using Fisher s z- test (Burns et al., 2003). I used one-half of the respondents data to calculate correlations between performance-only measures and overall trip satisfaction. I used the other onehalf to calculate correlations between difference scores and overall trip satisfaction. I used PROC CORR with the FISHER and SPEARMAN options in SAS v. 9.1 to conduct a one-tailed Fisher s z-test (SAS, 2003). I used Fisher s z-test to determine if

26 performance-only score and difference score correlations with overall trip satisfaction 17 differed significantly (Shavelson, 1996; Burns et al., 2003). I used alpha = 0.05 for significance testing throughout my study. I made no effort to contact individuals for a non-response survey who did not complete the pre-hunt questionnaire. I used a two-tailed Wilcoxon Rank Sum test to determine if expectations differed between those who completed pre-hunt and post-hunt questionnaires and those who only completed a pre-hunt questionnaire. To further check for possible non-response biases, I assumed each mailing wave probed deeper into the core of non-respondents (Filion, 1975; Choi, Dittion, & Matlock, 1992). Thus, using a two-tailed Spearman s rho, I correlated the mailing wave a participant responded to with variables related to: 1) overall trip satisfaction, 2) overall hunting experience satisfaction, 3) importance of hunting compared to other outdoor recreation activities, 4) importance of waterfowl hunting compared to other hunting activities, 5) days spent waterfowl hunting, 6) age, 7) annual household income, and 8) education. Using these methods, I would be able to identify possible non-response biases via significant correlations. My project was approved by the Mississippi State University Institutional Review Board (IRB) for the protection of human subjects (IRB docket number ). Results My study site was visited by 311 individuals, 187 (60.1%) of which agreed to participate in the study and completed the pre-hunt questionnaire. Of the 187 participants, 148 (79.1%) individuals responded to the post-hunt questionnaire. Five questionnaires were non-eligible (i.e., participants were less than 18 years old or

27 indicated the questionnaire was completed by another person) and removed from the 18 dataset, leaving 143 usable questionnaires from 182 eligible participants of the pre-hunt questionnaire for an effective mailing response rate of 78.6% (Dillman, 1978). Based on my correlation analysis of successive mailing waves, I found early respondents tended to be older (rho = , p = 0.004, n = 143), attained a higher level of education (rho = , p = 0.048, n = 143), and considered hunting more important than other outdoor recreation activities (rho = 0.239, p = 0.004, n = 141). I did not detect any statistically significant correlations for: 1) overall trip satisfaction (rho = 0.149, p = 0.077, n = 142), 2) overall hunting experience satisfaction, (rho = 0.146, p = 0.083, n = 142), 3) importance of waterfowl hunting compared to other hunting activities (rho = , p = 0.370, n = 141), 4) days spent waterfowl hunting (rho = , p = 0.242, n = 143), or 5) annual household income (rho = 0.035, p = 0.686, n = 138). Further, I did not detect statistical differences in expectations between those who filled out a pre- and posthunt questionnaire and those who filled out only a pre-hunt questionnaire (0.085 p 0.894, z 1.725). Most respondents were White or Anglo (99.3%, n = 142) and male (97.9%, n = 139), with an average age of 44.5 (SE = 0.9, n = 143) years. Respondents had a median gross annual household income of $100,000 and above and 86.0% (n =123) had some college or graduate level education. Survey participants had an average of 15.3 (SE = 1.3, n = 138) years of waterfowl hunting experience. Hunting was rated as the most important outdoor recreation activity for most (51.8%, n = 73) participants. Waterfowl hunting was rated as the most important hunting activity by 25.5% (n = 36) of

28 participants and as the second most important hunting activity by 38.3% (n = 54) of 19 participants. Respondents hunted an average of 3.2 (SE = 0.4, n = 141) days in Arkansas and 6.3 (SE = 0.9, n = 143) days outside of Arkansas during the waterfowl hunting season. For the single item measuring satisfaction with the overall trip, 90.8% (n = 129) of the respondents reported being either very satisfied or extremely satisfied, with an average score of 4.4 (SE = 0.1, n = 142). Positive disconfirmation occurred for 28%, confirmation occurred for 40%, and negative disconfirmation occurred for 32% of the 25 items measured (Table 2.1). I included negative difference scores as confirmed expectations because their 95% confidence intervals overlapped zero. There were differences between the difference scores reported in Table 2.1 and a difference score derived by subtracting the average performance from the average expectation because of rounding. Although participants did not harvest as many mallards (Anas platyrhynchos) as they expected, items related to skill testing, working ducks (calling at ducks circling decoys), seeing wildlife, and learning had greater negative disconfirmation. Thus, I rejected the hypothesis that negative disconfirmation would be greatest for harvestrelated variables. When I correlated performance-only items with satisfaction for the overall trip, all items were significantly related (0.001 p 0.015, rho 0.595; Table 2.2). When I correlated difference scores with overall trip satisfaction, 52% of items were significantly related (0.001 p 0.045, rho 0.451; Table 2.3). When I ranked items by correlation strength for difference scores and performance-only scores, the

29 20 strongest related items varied between measurement techniques. However, the variables of I was placed in the best available hunting location, I worked a lot of ducks, and I harvested a sufficient number of mallards were among the most significantly related (p 0.001) items to overall satisfaction for both methods. Because both of the harvestrelated items correlated significantly with overall satisfaction, I accepted the hypothesis that harvest-related variables would be significant predictors of overall satisfaction. Furthermore, I accepted the hypothesis that satisfaction would be influenced positively by fulfilled expectations because of the positive correlation with overall satisfaction for all significantly related items in Table 2.3. Performance-only measures correlated better with overall trip satisfaction than difference scores for 84% of the items. However, I detected significant differences for 40% of the item correlations with Fisher s z-test (0.001 p 0.047, z 3.274; Table 2.4). Thus, I rejected the hypothesis that performance-only measures would correlate better with overall satisfaction than difference scores for most items examined. Discussion and Implications I used methods typically used by customer satisfaction researchers in a context where items related to both customer service and hunting were potentially important to the overall satisfaction of clientele. Although some items with positive disconfirmation correlated relatively low with satisfaction, they may have had stronger influence on satisfaction ratings if expectations were not met. Therefore, managers should be cognitive to at least maintain standards for all strong correlates with trip satisfaction. Cohen (1988) suggested a correlation of 0.3 has a medium effect size and would be

30 21 perceptible to the naked eye of a sensitive observer. Thus, correlations of 0.3 and greater may offer a sufficient level of relatedness to warrant management attention. However, my study results suggest managerial efforts would best be focused on items that are strong correlates with satisfaction and have negative disconfirmation. For example, traditional performance-only measures would suggest placing clientele in the best available hunting location, placing them in a pristine environment, or providing a variety of entertainment options other than hunting would best satisfy customers (Table 2.2). However, expectations were met for being placed in the best available hunting location and having a variety of entertainment options other than hunting (Table 2.1). Alternatively, a manager could concentrate efforts to allow hunters to be in a pristine environment, test their skills, or work more ducks to have the best opportunity to improve overall satisfaction ratings. Nonetheless, managers may not have explicit control over these trip aspects and may only be able to improve these opportunities indirectly via habitat and hunting pressure management. Although using performance measures allows for similar management recommendations to be made, difference scores offer a greater breadth of understanding to satisfying clientele (Crompton & Love, 1995). Several explanations exist as to why items did not rank by correlation more similarly between the two measurement techniques. First, the relatively small number of participants in my study (n = 143) could be a factor; however, Burns et al. (2003) had many more respondents (n = 2,933) and correlations of their items were not similarly ranked either. Second, difference scores could rank dissimilarly because the added dimensionality they offer may allow for the true order of importance to be displayed via

31 correlation ranking. Third, Peter, Churchill, & Brown (1993) stated difference scores 22 may produce spurious correlations because difference scores are not unique from the components of which they are made. However, their alternative is not as likely in my study because of the split half design in comparing correlations. Finally, Parasuraman et al. (1994) offered shared method variance as a reason why performance-only measures correlated better with an overall evaluation of service quality than difference scores. Perhaps if difference scores were correlated with an expected level of overall satisfaction minus a perceived level of overall satisfaction, rankings would be more similar to the performance-only items correlated with a perceived level of overall satisfaction used in my study. However, the validity of using a difference score for overall satisfaction is unknown. Despite differences in correlation rankings, my study demonstrated comparable correlations for difference scores and performance-only measures for most items where other research did not (e.g., Burns et al., 2003). These findings could have been a result of my using the difference between will expectations and a measurement of performance for calculating difference scores, whereas Burns et al. (2003) used importance of an item and satisfaction of performance to calculate their difference scores. Niedrich et al. (2005) demonstrated will expectations had greater predictive ability than other types of expectancies. I also measured expectations prior to the recreational experience rather than after the experience as Burns et al. (2003). Another possible reason for better performance of difference scores in my study may relate to my use of Spearman s rho instead of Pearson s r. When using ordinal data, Spearman s rho is the

32 23 more appropriate analysis (Schlotzhauer & Littell, 1997). Finally, Burns et al. (2003) had a larger sample size than me and had greater statistical power for detecting differences (Cohen, 1988). I do not believe my follow-up survey suffered from non-response bias for the variables examined in my study, because the significantly correlated demographic characteristics are relatively poor predictors of attitudes (Manning, 1999) and no directional differences were found for satisfaction items. However, because of the high refusal rate for completing the pre-hunt questionnaire, I cannot be certain my sample was representative of the population studied. Choi et al. (1992) suggested it was not possible to determine if study populations were homogenous before surveys were conducted and data from non-respondents were difficult to obtain. In my study, it was not possible to get data from individuals who refused to complete the pre-hunt questionnaire; thus, I was unable to compare homogeneity between survey participants and individuals who refused to participate. The disconfirmation of expectations has been studied in great detail in the consumer satisfaction research, but, except for some earlier research (i.e., Decker et al., 1980), has been largely neglected by hunting satisfaction researchers. Although use of difference scores has opponents in the marketing literature (i.e., Cronin & Taylor, 1992; Peter et al., 1993), measuring the disconfirmation of expectations can offer valuable insight to wildlife managers wishing to gauge hunter satisfaction. Difference scores could be attained relatively easily on public hunting areas that use a lottery type system to regulate access. By mailing a pre-hunt questionnaire to gauge the expectations of hunters

33 drawn for a hunt, researchers could overcome the potential bias on expectations in my 24 study wherein questionnaires were distributed on site. If difference scores are not desirable to the researcher, disconfirmation also may be measured by allowing recreationists to evaluate their disconfirmation subjectively on a post-experience measurement scale as did Oliver (1980). However, more recent research has shown evaluating disconfirmation on a post-experience scale allowed participants to assimilate different disconfirmations into a similar construct, thus one may not be able to differentiate among the different types of expectations (Niedrich et al., 2005). Further, wildlife managers may benefit from research concentrating on the disconfirmation of different types of expectations. For example, Decker et al. (1980) used the difference between hunters perceptions of an ideal hunt and an actual hunting experience to establish hunter satisfaction management priorities. Whereas my study focused on the disconfirmation of predictive expectations, other research could examine normative expectations to improve quality hunting areas. My use of 95% confidence intervals to assign items to disconfirmation type offers another way to prioritize management efforts beyond simple ranking of difference scores, as done in earlier work (Parasuraman et al., 1988). Furthermore, researchers wishing to improve hunter satisfaction models could possibly look to the expectancy disconfirmation paradigm to gain greater explanatory power in satisfaction models. References Boulding, W., Kalra, A., Staelin, R., & Zeithaml, V. A. (1993). A dynamic process model of service quality: From expectations to behavioral intentions. Journal of Marketing Research, 30, 7-27.

34 Burns, R. C., Graefe, A. R., & Absher, J. D. (2003). Alternate measurement approaches to recreational customer satisfaction: Satisfaction - only versus gap scores. Leisure Sciences, 25, Choi, S., Ditton, R. B., & Matlock, G. C. (1992). Homogeneity across mail survey waves: A replicated study. Journal of Leisure Research, 24, Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates. Crompton, J. L., & Love, L. L. (1995). The predictive validity of alternative approaches to evaluating quality of a festival. Journal of Travel Research, 34, Cronin, J. J., Jr., & Taylor, S. A. (1992). Measuring service quality: A reexamination and extension. Journal of Marketing, 56, Cronin, J. J., Jr., & Taylor, S. A. (1994). SERVPERF versus SERVQUAL: Reconciling performance-based and perceptions-minus-expectations measurement of service quality. Journal of Marketing, 58, Decker, D. J., Brown, T. J., & Gutierrez, R. J. (1980). Further insights into the multiplesatisfactions approach for hunter management. Wildlife Society Bulletin, 8, Dillman, D. A. (1978). Mail and telephone surveys: The total design method. New York, NY: John Wiley & Sons, Inc. Ditton, R. B., Graefe, A. R., & Fedler, A. J. (1981). Recreational satisfaction at Buffalo National River: Some measurement concerns. USDA Forest Service General Technical Report: University of Minnesota Agricultural Experiment Station: St. Paul, NC. Filion, F. L. (1975). Estimating bias due to nonresponse in mail surveys. Public Opinion Quarterly, 39, Frey, S. N., Conover, M. R., Borgo, J. S., & Messmer, T. A. (2003). Factors influencing pheasant hunter harvest and satisfaction. Human Dimensions of Wildlife, 8, Gigliotti, L. M. (2000). A classification scheme to better understand satisfaction of Black Hills deer hunters: The role of harvest success. Human Dimensions of Wildlife, 5,